Multi-Omics Study Identifies Key Genes Regulating Gestation Length in Large White Pigs

A comprehensive multi-omics investigation published in Genetics Selection Evolution has identified GNAZ and SLC5A4 as core functional candidate genes governing gestation length across commercial populations of Large White pigs. Gestation length represents an essential component of reproductive efficiency and maternal productivity in swine breeding, exerting direct phenotypic effects on piglet birth weight, litter size, and postnatal survival rates.

The study cohort comprised 22,783 complete farrowing and reproductive records gathered between 2019 and 2024 from 9,057 sows reared across five specialized domestic breeding facilities in China. These experimental herds represented five distinct founding genetic lines originally established from Danish, French, Dutch Topigs, American, and Pig Improvement Company strains, with each strain maintained on an independent farm to prevent inter-line gene flow. Specifically, the analyzed cohorts included 2,574 sows from the Pig Improvement Company line, 1,238 sows from the Dutch Topigs line, 2,406 sows from the Danish population, 898 sows from the French strain, and 1,941 sows from the American lineage. Gestation length was precisely recorded as the continuous chronological time interval in days elapsed between successful artificial insemination and final piglet expulsion. Across the five examined strains, descriptive statistics revealed notable phenotypic variation in mean gestation durations.

To estimate additive genetic components and total phenotypic variability, the investigators applied a linear repeatability animal model implemented through the specialized variance estimation software DMU v6.0. The statistical framework evaluated adjusted phenotypic vectors by incorporating fixed effects for year-season groupings and parity classes alongside random service boar effects, maternal permanent environmental effects, additive genomic relationship matrices, and residual errors. Parity was classified into standardized sequential categories, while seasonal components were partitioned into calendar quarters spanning spring, summer, autumn, and winter. The resulting heritability estimates spanned from 0.17 to 0.37 across the individual breeding strains, confirming that porcine gestation length functions as a moderately heritable polygenic trait influenced substantially by maternal genomic variation. The highest single-population heritability was observed in the French Large White herd at 0.37, while heritability estimates for the American Large White cohorts settled at a consistent value of 0.30 regardless of which genotyping assay had been deployed. These quantitative findings demonstrate sufficient additive genetic variance to support direct genomic selection strategies aimed at optimizing farrowing schedules.

Genotyping across the study populations utilized multiple commercially validated arrays mapped directly to the Sus scrofa 11.1 pig reference genome assembly. Sows from the Pig Improvement Company population were assayed using the GeneSeek Genomic Profiler Porcine 50K bead chip, whereas the Topigs, Danish, and French cohorts were genotyped using the Compass Pig 50K Plus BeadChip. Within the American population, 991 animals were processed on the Compass Pig 50K Plus BeadChip, while an additional 950 animals were evaluated via a specialized 50K liquid-phase gene chip. Stringent quality control filters applied through PLINK v1.9 excluded animal samples exhibiting overall genotype call rates below 90 percent. Autosomal markers were subsequently filtered out if minor allele frequencies dropped below 0.05 or missingness thresholds exceeded 0.10. To elevate marker resolution to sequencing standards, the researchers utilized Beagle v5.2 against a multi-breed composite reference panel containing 1,662 whole-genome sequenced pigs sourced from PigGTEx and Suhuai cohorts, yielding imputation accuracy exceeding a 97.33 percent concordance rate and generating between 4.27 million and 6.36 million markers per herd.

Genome-wide association studies were conducted independently within each population on adjusted phenotype scores combining estimated breeding values and model residuals using LDAK v5.2. To safeguard statistical validity against spurious associations, independent principal component analyses performed in PLINK identified internal population stratification, prompting the inclusion of the top five principal components as fixed covariates within the linear models. Association testing deployed a leave-one-chromosome-out modeling strategy that prevents proximal candidate markers on the test chromosome from deflating overall test statistics while maximizing computational efficiency. Within the array-level analyses, 62 significant single nucleotide polymorphisms were identified in the Pig Improvement Company cohort, led by marker rs80812908 on chromosome 7, which accounted for 1.0 percent of phenotypic variance with a p-value of 4.66E-08. Imputed sequencing-level analyses dramatically magnified association signals, pinpointing 1,020 genome-wide significant variants in the Pig Improvement Company cohort on chromosomes 4 and 7, alongside 2,163 significant variants across chromosomes 2 and 14 within the Danish population.

The cross-population meta-analysis successfully identified 3,134 single nucleotide polymorphisms surpassing the stringent genome-wide threshold of 2.04E-07, with significant association peaks mapping to Sus scrofa chromosomes 2, 3, 7, 9, 12, and 14. Meta-analytic aggregation substantially strengthened the statistical significance of leading quantitative trait loci compared with single-breed scans. On chromosome 2, the leading p-value dropped from 6.95E-09 to 9.88E-10, while the chromosome 7 peak dropped from 3.20E-10 to 2.16E-11. On chromosome 9, an association interval spanning 127.29 to 128.61 megabases that was previously detected at suggestive levels in single-herd analyses attained a definitive meta-analytic p-value of 8.43E-15 at position 127,495,052 bp. Furthermore, the cross-population meta-analysis revealed an entirely novel significant marker, rs330354247, located at position 58,826,881 bp on chromosome 3, demonstrating the power of cross-population sample pooling.

To transition from broad chromosomal association intervals to discrete functional targets, the authors applied the Bayesian fine-mapping algorithm CAVIARBF v0.2.1 to the meta-analysis summary data. The statistical model integrated marginal association test statistics with local linkage disequilibrium correlation structures across a one-megabase window flanking each leading variant, computing posterior inclusion probabilities to isolate the smallest credible variant set containing the causal mutation with 95 percent probability. This fine-mapping approach successfully narrowed candidate intervals on chromosome 7 to two distinct segments spanning 10,460,892 to 10,737,692 base pairs and 31,384,086 to 31,384,333 base pairs, the latter representing an ultra-refined region of only 247 base pairs. Additionally, the quantitative trait locus on chromosome 9 was narrowed to an 883-kilobase window between 127,322,114 and 128,205,369 base pairs, the chromosome 12 locus was delimited between 24,640,681 and 25,479,574 base pairs, and the primary chromosome 14 locus was refined to an interval between 48,350,921 and 49,415,663 base pairs.

The study then incorporated functional transcriptomic profiling by conducting transcriptome-wide association studies across 34 distinct porcine tissue expression panels cataloged in the PigGTEx database. Implementing the FUSION model, the researchers correlated predicted genetically regulated gene expression levels with gestation length phenotypes, focusing specifically on endocrine and reproductive tissues relevant to embryonic survival. Within the fine-mapped chromosome 14 interval spanning 48.25 to 49.02 megabases, two candidate genes exhibited genome-wide significance: GNAZ and SLC5A4. Genetically regulated expression of GNAZ in the pituitary gland, spleen, duodenum, and adipose tissues correlated significantly with gestation length variations, achieving p-values below 7.09E-07. Similarly, expression levels of SLC5A4 within embryonic tissue, cartilage, spleen, and pituitary gland demonstrated significant associations with gestation length, with p-values falling below 5.15E-06.

To establish the molecular regulatory mechanisms operating across the chromosome 14 locus, the investigators mapped multi-omics epigenetic marks utilizing the IAnimal comparative genomics database. Chromatin accessibility assays via ATAC-seq across porcine hindbrain tissues identified prominent open chromatin peaks directly adjacent to the GNAZ transcriptional start site. Parallel chromatin immunoprecipitation sequencing revealed that active histone modifications, specifically H3K27ac marks associated with active enhancers and promoters, colocalized directly with these accessible chromatin segments in brain tissue. Conversely, open chromatin marks were absent across available datasets for SLC5A4, indicating distinct transcriptional regulation. Cross-species transcriptomic comparisons between 26 porcine tissues from PigGTEx and 25 human tissues from GTEx established that both GNAZ and SLC5A4 maintain evolutionary conserved expression hierarchies, with both genes displaying pronounced transcriptional abundance in human and porcine pituitary gland tissues.

The biological convergence of epigenetic accessibility, transcriptomic associations, and statistical fine-mapping highlights distinct mechanistic roles for GNAZ and SLC5A4 during gestation. Furthermore, GNAZ acts as an intracellular regulator of calcium signaling and protein kinase C activity within immune cells, mechanisms thought to sustain maternal immune tolerance toward developing fetuses across pregnancy. Nonetheless, the identified genomic regions and candidate markers provide molecular targets for marker-assisted selection programs seeking to optimize porcine gestation length and reduce reproductive losses in commercial swine herds.

Subject of Research: Identification of genetic regulators of gestation length in Large White pigs using integrative multi-omics analysis

Article Title: Integrative multi-omics analysis identifies GNAZ and SLC5A4 as critical regulators of gestation length in Large White pigs

Article References: Zhou, J., Zhao, Q., Qian, X., Zeng, T., Xu, L., Liu, Q., Yin, Y., Ma, J., Yu, J., Jiang, Z., Zhao, G., Huang, R., & Li, P. (2026). Integrative multi-omics analysis identifies GNAZ and SLC5A4 as critical regulators of gestation length in Large White pigs. Genetics Selection Evolution, 58(1), Article 41. https://doi.org/10.1186/s12711-026-01058-5

Image Credits: AI Generated

DOI: 10.1186/s12711-026-01058-5

Keywords: Large White pigs, gestation length, GNAZ, SLC5A4, genome-wide association study, GWAS meta-analysis, Bayesian fine-mapping, transcriptome-wide association study, phenome-wide association study, epigenetic profiling, porcine reproduction, quantitative trait loci, genotype imputation, PigGTEx, heritability, swine genomics